790 karma · joined November 7, 2014
One non-obvious aspect is that the function relies on using Python's big ints, even for pretty small values of n. You are calculating `2^(n(n+1))` for the numerator, so if n=8 you already get `2^72`, which you can't fit in a 64-bit int. This is even though the answer itself is very small: `F_8=21`.
This makes sense, because you are essentially packing all the previous Fibonnaci numbers into a bit string, so you need room for all of them without overlapping and that bit-string is going to necessarily be long.
`The biggest loss of all, though, is a necessary one: the myth that anything but a for-profit corporation is the right way to organize a company.`
I don't see how anything that happened this weekend leads to this conclusion. Yes, it seems likely that the board's actions will result in an OpenAI with much smaller revenue and consumer traction. But the whole reason for setting up OpenAI as a non-profit was precisely ensuring that those were not the overriding goals of the company.
The only conclusion that seems warranted is that "the myth that anything but a for-profit corporation is the right way to organize a for-profit company.", but that is pretty obvious.
(The other passage was the one about one of the characters using a computer while in prison and setting up a profusion of xterms scrolling through random files because he suspected he was being spied on using van Eck phreaking.)
If the choice was between those two extremes, I'd take the authoritarian nightmare.
Extinction is forever, dictatorships end.
It would be better if we didn't need to do this, but I don't see a less intrusive way.
Of course, if you don't think there is a real existential threat then it's pointless and bad to do this, it all depends on that premise.
My take is that if you remove nuclear/biological war out of the picture somehow, climate change is not an existential risk. If you remove the latter then the former is still an existential risk (and there are unfortunately a lot of other possible sources of geopolitical instability). So the fundamental source of risk if the former. But it's a matter of opinion.
Conventional or chemical warfare, even on a global scale, are definitely not existential risks though. And like I said, probably not even nuclear. Biological... that I could see leading to extinction.
Still, like I said somewhere else, even all out nuclear war is unlikely to lead to human extinction. It could push us back to the neolithic in some scenarios, but even then there is some disagreement about what would happen.
Of course, even in the most optimistic scenario it would be really bad and we should do everything we can to avoid it - that goes without saying.
I would still bet against going back to 19th century. Worst case scenario for me is hundreds of millions dying and a big regression in global standards of living, but keeping our level of civilization. Which is horrible, it would be the worst thing that ever happened.
I am just saying it's not an extinction risk.
But according to your link, IPCC says it's not a plausible scenario in the medium term. And I'd say that even 8 degrees of warming wouldn't be enough for extinction or end of human civilization. But there it could be double digit percentage of human population dying.
It was not very fair from civil war already, with the aborted coup. Could you imagine if the Wagner group became a nuclear power??
I don't think IUPAC (International Union of Pure and Applied Chemistry) has come out with views on impact of global warming :)
Human extinction via climate change is off the table.
The end of human civilization? Depends on what one means by it. If it means we'll go back to being hunter gatherers it's ridiculous. Some civilizational wide regression is not impossible, but even the direst IUPAC scenario projects only a one digit drop in global GDP (of course, that's the average, it will be much worse in some regions).
There is some risk that over time people will be able to do more with less, and eventually you can train an AGI with a few dozen A100s. If that happens I agree there's nothing you can do, but until then there is a chance.
First nuclear weapons were used in 1945, almost 80 years ago. Today, out of ~ 200 countries in the world only 9 have nukes. Of those, only 4 did it after Non-proliferation of Nuclear Weapons Treaty. South Africa had them and gave them up under international pressure. Ukraine and other former soviet republics voluntarily gave them up after the breakup of the USSR.
So, yes. Unfortunately we still have nukes and there are way too many of them, but it's not like non-proliferation efforts achieved nothing. Iran, a large country with a lot of resources, has been trying to get them a for a while with no luck for instance.
And for frontier AI research you don't just need matrix multiplication, you need a lot of it. There's less than a handful of semiconductors fabs globally that can make the necessary hardware. And if you can't get to TSMC for some reason, you can get to ASML, NVDIA, etc.
Nuclear war, if the worst nuclear winter predictions are correct, could be a lot worse but there would still be some survivors.
Unaligned ASI though, could actually make us extinct.
[1] https://blog.cloudflare.com/scalable-machine-learning-at-clo...
In some cases, like vectors and axial vectors, it is less obvious because (in 3 spatial dimensions), they actually have the same number of components. But still, if you add them together you get something that does not transform in a sensible way under reflections.
However, regarding the Dirac delta function an other generalized functions, even though they were introduced by physicists (I think Heaviside was an early proponent) in a hand wavy fashion, they were later put in rigorous mathematical footing in the theory of distributions. Distributions are nowadays used by mathematicians without any hesitation.
This also not the first instance of something like this, and it won't be the last. Physicists have come up with ad hoc methods that work, but they can't justify why with rigor. Some time later mathematicians formalize it and it becomes part of their tool set.
Looking closer, it's not as bad as what I had in mind. Association order only depends on type, not value (which is what I meant by data). This means you can figure it out at compile time, and actually makes some sense if you want to be able to use standard mathematical notation like they do, which has operators with different association orders.
- Order of operations depends on number of terms in more complicated manner.
- Order of operations depends on data? This would make compiling the language basically impossible, but you could probably get away with it in an interpreter.
- Parenthesis have side effects, in addition to changing order of operations.
Likewise. It's disappointingly rare to be able to have a civil discussion with someone you have very different opinions from.
> The gap between professionals and hobbyists is the same between professionals and customers, its the main reason of trustee obligations / market regulations. So something about this issue is not fine.
> My point is, privileged market access leads to growing power asymmetries, which a newcomer can never compensate with genius abilities alone. This kind of market capture is similar to institutionalized corruption btw.
And my point is that you are holding trading to a different standard. Is it a problem that Apple has so many resources that a hobbyist has no chance of competing with at making smartphones? Or Google and search engines, Chipotle and burritos, etc?
Of course not. Those companies have earned their market shares by investing in the right places, innovating, good strategy, etc. Same is true for HFT firms.
There would be a problem if either Apple or Citadel had a monopoly or was breaking the law. If there's proof that's the case, we have courts and regulators to step in.
> I replied to your stronghold statements, didnt i?
> Now, is HFT just a client side approach of front running? Which is an insider privilege and illegal.
According to the legal definition, front running is when you have orders you are executing for clients and you trade ahead of them on your own account. HFTs do not have client orders, ergo they cannot possibly front run anything.
If you use some expanded definition of front running, then it depends on the details of the definition. For some sufficiently broad definitions, probably yes. But then it's of course not illegal and possibly not even bad.
They are only "pushing out" other liquidity providers. If you are a medium/long term investor you are not competing with HFTs. Yes, if you want to be day trader HFTs might make your life harder, but I think from society's point of view that's fine.
> The only way to seriously compete with such HFTs, is to do it as they do, which requires the same physical accesses to the markets. Previously, the access to valuable (insider) information was a privilege for slow traders and a way to get an edge. Speed and algorithms are imo just a new privilege, enabled by technology.
> The simple reason why insider trading is illegal, because it puts other marketeers at a disadvantage, they have no means to compensate. The same is true for HFT. You cannot rent the rack next to the markets machines when its already occupied.
There is no principle that all market participants must been on equal footing. That would be impossible to enforce, and counterproductive to boot. If you are a slow (non HFT) trader at a big fund or bank you also have a lot of advantages over a guy doing it from home: better financing rates, more market access, a whole infrastructure to provide you with information and analysis, etc. Should that be made illegal as well?
What about if you are just a better trader? Should you be somehow given a handicap so others can compete with you?
As I said in another reply, in every field of human activity, professionals are advantaged over hobbyists, and that's fine.
Liquidity provision is more than that. Without market makers/liquidity providers prices fluctuate a lot more because the balance of natural buyers/sellers varies over time. Liquidity provision is a shock absorber to those imbalances.
> But you are indirectly collecting economic rents by trading so fast that retail investors can't connect with each other, they always end up connecting with the HFT.
You are right that retail traders trade against HFTs most of the time. This is actually true for institutional investors as well, simply because HFTs provide almost all of the liquidity in stock market. I.e., it's not because HFT as are reacting to incoming orders faster (it is impossible to react to a marketable order before it trades by design), but because HFTs are the ones posting at the bid or ask.
For retail traders there's the extra step that most (all?) retail brokers have agreements with HFTs where they send them their retail orders before they hit the markets. The HFTs can opt to trade against the retail trader but only if they give them a better price than they could have gotten at the exchanges.
There is some unfounded misperception that this practice must be nefarious, based on the idea that the HFTs can only be doing this if they are pulling one over the retail traders. This is wrong, the reason HFTs like to trade against retail is that it is unlikely that a retail order is part of a large multi-million share meta-order that is going to push the price against the liquidity provider. This means that they can give retail traders lower spreads and still make money, both parties win. The party who loses in this are the institutional traders - since HFTs manage to segregate part of the retail flow, they end up using higher spreads for the rest of the flow.
> So in that sense you are collecting rent on the expensive high speed trading infrastructure by getting to the front of every queue.
This is stretching the definition of rent. If I own a well-run restaurant, am I collecting rent on expensive food preparation infrastructure that lets me prepare good food at a lower cost that my competitors?
> A brilliant/knowledgeable retail trader is unable to monetize their comparative advantage because they can't compete with the HFT operator on execution speeds.
They also have a hard time competing against traders working at big banks or hedge funds even if they don't use HFT infrastructure. In every field of human activity, professionals have big advantages over hobbyists, and that's fine.
Also note that if you are a retail trader buying stocks with a long/medium term outlook HFT is good for you (lower trading costs!). If you want to be day trader, HFT might make your life harder, but it's not clear to me that's a bad thing.
> The whole thing is built on FOMO, and creates the illusion of value by endlessly accelerating transaction speeds.
Transaction speeds are not endlessly accelerated. There's a misapprehension that HFTs are trading in and out of positions in a millisecond time scale. That's absurd, you'd never make money that way. They care about their latency down to the microsecond or hundreds of nano-seconds because that lets them cancel orders quickly when it looks like the market is going to move against them and avoid adverse selection, the bane of the liquidity provider.
> This is like arguing that people who drive way above the speed limit are creating value by arriving at their destination sooner (leaving more road space available) and encouraging others to drive fast (making the road transport system more efficient). Accidents that occur int he vicinity of speeding drivers' trajectory can be blamed on inferior driving standards.
For the faster drivers, there is clear downside: the increased accidents. For HFT, what is that?
> [...] The more it is automated and accelerated, the more it tends toward monopolistic concentration of capital in corporate entities and increases overall fragility of the system.
HFT firms require very little capital compared to things like big banks or large hedge funds. I don't have the numbers but I am sure their combined capital is tiny fraction of the rest of players in financial markets.
They don't hold large books ever, so if they go under, there is no fire sale that drives asset prices down and cause a crisis. They don't present a big systemic risk.
> Imho, all those statements are weak.
> Providing liquidity in markets is essential but markets worked long before HFT was a thing.
Correct, but the previous incarnation of liquid providers (mostly humans on stock exchange floors) was far more expensive and less efficient than HFT. This is why HFT was able to compete on price of liquidity provision (i.e., the spreads came down) and win.
These legacy firms did lose out from HFT (and are actually behind some of the anti HFT rhetoric out there) but I assume people in HFT see technological disruption as a good thing, even if bad for some incumbents.
The idea that retail traders or institutional investors are the ones that lost out from HFT lacks any evidence, AFAIK.
> FTH is just another way of going "meta" (for VC?) simply beeing faster. Until they are the market.
Your wording is a bit unclear, but if you are hinting that HFTs are operating at a loss until they capture sufficient market share, this is demonstrably false. HFTs make a lot of money (you can look up Virtu's financial statements for proof since it is a public company).
> The big question, as you too said is.
> > It is still possible that changing market structure or regulation [...] could be somehow beneficial for markets or society as a whole, but this is far from obvious.
> To me, it is rent seeking.
It would be if it was true. I admit it's a theoretical possibility, but I am yet to see a convincing argument of it.